AI LeakLake – Your AI Chat Leak Detector Real-time protection against accidentally exposing secrets in ChatGPT, Gemini, Claude, Perplexity, Grok, and every other LLM website. AI LeakLake scans everything you type or paste before it leaves your browser and instantly warns you if sensitive, dangerous, or regulated information is detected. What AI LeakLake watches for (90+ high-precision detectors running locally):

API keys & tokens
Private keys (SSH, PGP, RSA), JWTs, bearer tokens, passwords, database strings
PII & GDPR data (emails, phone numbers, SSNs, credit cards, passports, health records, children’s data, EU VAT numbers)
Crypto dangers (12/24-word seed phrases, wallet-drain phishing, fake airdrops, rug-pull talk)
Corporate secrets (messages, webhooks, S3 URLs, salaries, M&A details, roadmaps, customer lists, pricing strategy)
Illegal/harmful content (malware, drugs, weapons, child exploitation indicators, self-harm, scams)
Adult/explicit language and links
Source code snippets, Kubernetes secrets, internal domains, and more

Free Version: - Real-time detection and clear, non-intrusive alerts - Shows exactly what was found and why it’s risky - Runs 100% locally – nothing is ever uploaded or logged - Perfect for individuals who want awareness without automatic blocking Premium Version – Full Active Protection (coming soon): - Everything in Free, plus: - Automatic blocking – prevents the message from being sent when critical leaks are detected - Custom allow/deny rules (e.g., permit your own API keys, block everything else) - Team policies & central override - Detailed audit log of blocked items - Priority updates for new leak patterns Your data never leaves your browser – in both free and premium versions. No accounts, no telemetry, no cloud processing – ever. AI LeakLake – Because one leaked API key, seed phrase, or patient record is one too many. Stay aware with Free. Stay protected with Premium.

Latest version

Version 1.0.4

Older versions

Version 1.0.2

Version 1.0.1